{"schema_version":"newruntime-agent-readable-v0.2","type":"post","stable_id":"post:claude-cowork-recorded-skills","slug":"claude-cowork-recorded-skills","title":"Claude Cowork Turns Screen Recordings Into Reusable Skills","description":"Claude Cowork's recorded-skill flow makes workflow capture a first-party path from human demonstration to reusable agent capability.","retrieval_nugget":"Claude Cowork's recorded-skill flow makes workflow capture a first-party path from human demonstration to reusable agent capability. Claude Cowork is moving skill creation from written instructions toward observed work. The reported flow records a screen walkthrough while the user narrates the task, then turns that demonstration into a skill Claude can reuse later.","status":"published","published_at":"2026-07-23","updated_at":"2026-07-23","record_date":"2026-07-23","date_kind":"published_at","topics":["agent-skills","workflow-automation","agent-memory"],"source_urls":["https://x.com/claudeai/status/2079595988998554047","https://www.searchenginejournal.com/anthropics-claude-can-now-watch-a-video-and-learn-your-job/583053/"],"visuals":[{"id":"claude-recorded-skill-loop","kind":"editorial-diagram","role":"hero","src":"https://newruntime.com/images/posts/claude-recorded-skill-loop.webp","alt":"A six-stage workflow turns a narrated screen demonstration into an inspected, tested, versioned, and reusable agent skill with a failure loop back to review.","caption":"Recorded workflows are draft automation: record, extract, inspect, test, version, and only then reuse.","credit":"New Runtime synthesis from Claude Cowork's recorded-skill workflow","source_url":"https://x.com/claudeai/status/2079595988998554047","generated_with":"gemini-3.1-flash-image","width":1600,"height":900,"legend":[{"label":"Record","description":"Capture the human demonstration and narrated intent across the relevant work surfaces."},{"label":"Extract","description":"Turn the observed sequence into a draft procedure instead of preserving a raw transcript."},{"label":"Inspect","description":"Remove accidental clicks, hidden credentials, fragile UI assumptions, and outdated business rules."},{"label":"Test","description":"Run the cleaned procedure on a real bounded task and route failures back to inspection."},{"label":"Version","description":"Approve and store an inspectable skill package only after the behavior is reproducible."},{"label":"Reuse","description":"Call the versioned skill in later sessions while keeping the procedure editable."}]}],"routes":{"html":"https://newruntime.com/posts/claude-cowork-recorded-skills/","markdown":"https://newruntime.com/posts/claude-cowork-recorded-skills.md","json":"https://newruntime.com/posts/claude-cowork-recorded-skills.json"},"source_format":"markdown","next_reads":[{"type":"topic","path":"/topics/agent-memory/","reason":"Explore the agent memory topic hub.","url":"https://newruntime.com/topics/agent-memory/","title":"Agent memory - New Runtime","media_type":"text/html"},{"type":"related_material","path":"/posts/hermes-agent-learning-loop/","reason":"Shares agent memory and agent skills.","url":"https://newruntime.com/posts/hermes-agent-learning-loop/","title":"Hermes Agent Makes the Learning Loop Part of the Runtime","media_type":"text/html"},{"type":"related_material","path":"/posts/abbel-belief-state-memory/","reason":"Shares agent memory.","url":"https://newruntime.com/posts/abbel-belief-state-memory/","title":"ABBEL Treats Memory as an Explicit Belief State","media_type":"text/html"},{"type":"related_material","path":"/posts/openworker-local-desktop-coworker/","reason":"Shares workflow automation.","url":"https://newruntime.com/posts/openworker-local-desktop-coworker/","title":"OpenWorker Brings The Coworker Agent Back To The Desktop","media_type":"text/html"},{"type":"related_material","path":"/posts/mem0-context-wipe-memory-layer/","reason":"Shares agent memory.","url":"https://newruntime.com/posts/mem0-context-wipe-memory-layer/","title":"Mem0 Shows Agent Memory Should Not All Live in Context","media_type":"text/html"}]}